A video coding/decoding system builds implied reference frames from a plurality of reference frames developed during coding. Coded data of reference pictures are decoded and stored in a reference picture cache. An implied reference frame may be derived from a plurality of reference frames and may be stored in the reference picture cache. Thereafter, coding of new input data may proceed using the implied reference frame as a source of prediction. The method may be used to identify visual elements such as background elements that may persist in video over a longer period that can be captured by the system under conventional reference frames assembly and eviction. Implied reference frames may be built at both an encoder and a decoder to serve as sources of prediction.
Legal claims defining the scope of protection, as filed with the USPTO.
1. A video coder management method, comprising: decoding coded data of a plurality of reference pictures, the coded reference picture data having been transmitted in a channel between an encoder and a decoder, storing the decoded reference pictures in a reference picture cache, building a new, implied reference picture from a plurality of the stored reference pictures, and storing the newly built implied reference picture in the reference picture cache, wherein the method is executed at an encoder and the reference picture cache storing the reference pictures and implied reference picture is located at the encoder, transmitting a message to a decoder commanding the decoder to derive an implied reference frame, and predictively coding a new input frame with reference to the stored implied reference picture only after receipt of an acknowledgment to the message from the decoder.
2. The method of claim 1 , wherein the implied reference picture is derived without transmission of such through the channel.
3. The method of claim 1 , wherein implied reference picture data is derived from detected areas of commonality among the reference pictures.
4. The method of claim 1 , wherein implied reference picture data is derived from pixel blocks of the reference pictures identified as belonging to an image background.
5. The method of claim 4 , wherein pixel blocks are identified as belonging to the image background based on motion of the respective pixel blocks.
6. The method of claim 4 , wherein pixel blocks of a reference frame are identified as belonging to the image background based on a comparison of the pixel blocks' motion vectors to global motion of the reference picture.
7. The method of claim 4 , wherein pixel blocks of a reference frame are identified as belonging to the image background based on a correlation between the pixel blocks and spatially neighboring pixel blocks.
8. The method of claim 1 , further comprising predictively coding a new input frame with reference to the stored implied reference picture.
9. The method of claim 1 , further comprising decoding coded video data of a new frame with reference to the stored implied reference picture.
10. The method of claim 9 , wherein the coded video data includes an identifier of the implied reference frame as a prediction reference.
11. The method of claim 1 , wherein the message identifies long term reference frames to be used to derive the implied reference frame.
12. The method of claim 1 , wherein the message identifies reference frames to be used to derive the implied reference frame.
13. A video decoder management method, comprising: decoding coded data of a plurality of reference pictures, the coded reference picture data having been received on a channel from an encoder, storing the decoded reference pictures in a reference picture cache, deriving a new, implied reference picture from a plurality of the stored reference pictures, and storing the newly built implied reference picture in the reference picture cache, wherein the deriving is performed in response to a message received from an encoder commanding the decoder to derive an implied reference frame, after the implied reference frame is derived, transmitting an acknowledgment of the message to the encoder, wherein the method is executed at a decoder and the reference picture cache storing the reference pictures and implied reference picture is located at the decoder.
14. The method of claim 13 , wherein the implied reference picture is derived without receiving such from the channel.
15. The method of claim 13 , wherein implied reference picture data is derived from detected areas of commonality among the reference pictures.
16. The method of claim 13 , wherein implied reference picture data is derived from pixel blocks of the reference pictures identified as belonging to an image background.
17. The method of claim 16 , wherein pixel blocks are identified as belonging to the image background based on motion of the respective pixel blocks.
18. The method of claim 16 , wherein pixel blocks of a reference frame are identified as belonging to the image background based on a comparison of the pixel blocks' motion vectors to global motion of the reference picture.
19. The method of claim 16 , wherein pixel blocks of a reference frame are identified as belonging to the image background based on a correlation between the pixel blocks and spatially neighboring pixel blocks.
20. The method of claim 13 , further comprising decoding coded video data of a new frame with reference to the stored implied reference picture.
21. The method of claim 20 , wherein the coded video data includes an identifier of the implied reference frame as a prediction reference.
22. The method of claim 13 , wherein the message identifies long term reference frames to be used to derive the implied reference frame.
23. The method of claim 13 , wherein the message identifies reference frames to be used to derive the implied reference frame.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
January 7, 2011
September 23, 2014
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.